GENERALIZED STOCHASTIC GRADIENT LEARNING

George W Evans; Seppo Honkapohja; Noah Williams

doi:10.1111/j.1468-2354.2009.00578.x

GENERALIZED STOCHASTIC GRADIENT LEARNING

George W Evans, Seppo Honkapohja, Noah Williams

Research output: Contribution to journal › Article › peer-review

Abstract

We study the properties of the generalized stochastic gradient (GSG) learning in forward-looking models. GSG algorithms are a natural and convenient way to model learning when agents allow for parameter drift or robustness to parameter uncertainty in their beliefs. The conditions for convergence of GSG learning to a rational expectations equilibrium are distinct from but related to the well-known stability conditions for least squares learning.

Original language	English
Pages (from-to)	237-262
Number of pages	26
Journal	International Economic Review
Volume	51
Issue number	1
DOIs	https://doi.org/10.1111/j.1468-2354.2009.00578.x
Publication status	Published - Feb 2010

Keywords

MONETARY-POLICY
NASH INFLATION
STABILITY
EXPECTATIONS
CONVERGENCE
FRAMEWORK
BELIEFS
RULES
MODEL

Access to Document

10.1111/j.1468-2354.2009.00578.x

Cite this

@article{f14e4b7d9601406caf4c910036fc5da4,

title = "GENERALIZED STOCHASTIC GRADIENT LEARNING",

abstract = "We study the properties of the generalized stochastic gradient (GSG) learning in forward-looking models. GSG algorithms are a natural and convenient way to model learning when agents allow for parameter drift or robustness to parameter uncertainty in their beliefs. The conditions for convergence of GSG learning to a rational expectations equilibrium are distinct from but related to the well-known stability conditions for least squares learning.",

keywords = "MONETARY-POLICY, NASH INFLATION, STABILITY, EXPECTATIONS, CONVERGENCE, FRAMEWORK, BELIEFS, RULES, MODEL",

author = "Evans, \{George W\} and Seppo Honkapohja and Noah Williams",

year = "2010",

month = feb,

doi = "10.1111/j.1468-2354.2009.00578.x",

language = "English",

volume = "51",

pages = "237--262",

journal = "International Economic Review",

issn = "0020-6598",

publisher = "Wiley-Blackwell Publishing Ltd",

number = "1",

}

TY - JOUR

T1 - GENERALIZED STOCHASTIC GRADIENT LEARNING

AU - Evans, George W

AU - Honkapohja, Seppo

AU - Williams, Noah

PY - 2010/2

Y1 - 2010/2

N2 - We study the properties of the generalized stochastic gradient (GSG) learning in forward-looking models. GSG algorithms are a natural and convenient way to model learning when agents allow for parameter drift or robustness to parameter uncertainty in their beliefs. The conditions for convergence of GSG learning to a rational expectations equilibrium are distinct from but related to the well-known stability conditions for least squares learning.

AB - We study the properties of the generalized stochastic gradient (GSG) learning in forward-looking models. GSG algorithms are a natural and convenient way to model learning when agents allow for parameter drift or robustness to parameter uncertainty in their beliefs. The conditions for convergence of GSG learning to a rational expectations equilibrium are distinct from but related to the well-known stability conditions for least squares learning.

KW - MONETARY-POLICY

KW - NASH INFLATION

KW - STABILITY

KW - EXPECTATIONS

KW - CONVERGENCE

KW - FRAMEWORK

KW - BELIEFS

KW - RULES

KW - MODEL

UR - https://www.scopus.com/pages/publications/77950343648

U2 - 10.1111/j.1468-2354.2009.00578.x

DO - 10.1111/j.1468-2354.2009.00578.x

M3 - Article

SN - 0020-6598

VL - 51

SP - 237

EP - 262

JO - International Economic Review

JF - International Economic Review

IS - 1

ER -

GENERALIZED STOCHASTIC GRADIENT LEARNING

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this